NDVI data can be applied to estimate primary biomass by rbb25794

VIEWS: 270 PAGES: 20

									2.5.2. NDVI Application for Estimating End of Season Primary Biomass Production
NDVI data can be applied to estimate primary biomass production for the rangelands. This is a
calculation that more accurately estimates herbaceous biomass as opposed to woody biomass.
The primary biomass production in 1998 and 2000 is shown in fig 2.5. This data is used to
assess land carrying capacity for wildlife and livestock and grazing pressure for different range
units. It is a good management tool in drought mitigation.




Fig 2.5. Primary biomass production estimates derived from NDVI data in Kg/ha for the long rains in
1998 and 2000 respectively.


2.5.3 Drought Monitoring at District Scale

It is possible to illustrate the effect of the drought at any area of interest. For example, Isiolo
district was chosen to illustrate the use of NDVI to monitor the progression of seasons using
archival information. This can form the basis of an effective early warning system.
Comparisons of satellite data allows the detection of change and provide appropriate threshold
value associated with actual change. A series of multi-date imagery can provide information
on trends in progression of seasonal development. For example, if a substantive data archive is
available, statistical comparison of current events can be made to highlight anomalies in
seasonal conditions as illustrated in Figure 2.6.




                                                 40
                                    1 - 10 M ar 19 9 5 (n orm al)                            1 - 10 M ar 19 9 6 (d rou gh t)




                             1 - 10 M ar 19 9 7 (h igh rain fal)
                                                               l                                                    -
                                                                                              1 - 10 M ar 19 9 8 (El Nin o


                  Fig. 2.6. NOAA NDVI images showing rainfall anomalies at the start of long rains
                  (March) between 1995 and 1998 in Isiolo district.


Laikipia, Samburu and Isiolo districts (fig 2.7) form an important northern ASAL ecosystem
for both wildlife and livestock depending on availability of pasture and water. Animals
normally move southwards towards Laikipia district during drought resulting in conflicts over
limited pastures and watering points. In the past, such movements formed part of the coping
mechanisms, but with increased settlements and commercial ranches in Laikipia such
movements pose a great management challenge.


                                                                                                                    3 MAY 0 0
               3 MAY 9 8                                               3 MAY 9 9




                                                                                                          SAMBURU
     SAMBURU                                                 SAMBURU                                                            ISIOLO
                           ISIOLO                                                   ISIOLO




                                                                                                     LAIKIPIA
LAIKIPIA                                                LAIKIPIA




Fig. 2.7. SPOT 4 NDVI images showing rainfall variability over the same period in 1998, 1999 and
2000. The country received the El Nino rains (1997-98) that resulted in heavy flooding, which was
followed by one of the worst droughts in 1999-2000. This steep variation in a short span calls for
preparedness in mitigating impacts caused by extreme environmental conditions.




                                                                                   41
 Fig.2.8. Primary biomass production estimates in kg/ha derived from NDVI data for Laikipia,
 Samburu and Isiolo districts during the El Nino rains and the year 2000 drought.


2.6 ANALYSIS OF NDVI ANOMALIES
The following graphs show the deviation from normal using historical NOAA NDVI data for
selected districts in Kenya chosen from the high potential areas, the southern and the northern
rangelands during 1982-2004. The representation of the diverse agro ecological zones of the
country helps in understanding the impacts in the different zones of the country. Years in
which drought was persistent is shown by troughs below normal, which was the zero axis of
the NDVI anomalies. The critical drought years in each district are highlighted in gray shade.
The droughts in most districts show a cyclic pattern with 1984 reporting drought in most
districts followed by minor ones in 1994 and the severest recorded in the year 2000.
          The southern rangeland districts of Kajiado and Narok and northern rangeland districts
of Isiolo and Laikipia were selected. Aerial surveys were undertaken in the districts to monitor
livestock and wildlife populations at the peak of the drought. Nakuru and Muranga were
selected to represent the high potential districts. In the southern rangeland the analysis of the
NDVI anomalies revealed a cyclic recurrence of drought episodes from 1982 to 2000. Kajiado
district (Figure 2.9) recorded a major drought during the 1982-1984 period. Other droughts in
the district occurred in 1992, 1994, 1996 and 1997. The severest drought in the district
occurred in 2000 recording the highest NDVI anomaly of more than –2. This result was
corroborated from the field report of the district showing 1984 and 2000 drought as the worst
in the district in recent years. Although the 2000 drought appears to have been very severe
from the NDVI anomalies results, people interviewed perceived 1984 drought to be the worst.
This can be partly explained from the fact that 1982-1984 drought was persistent without
allowing time for recovery hence the severity of the impacts at the time in the district. The
1999-2000 lasted a short time although recording very high NDVI anomaly.

      3


      2
                                                42


      1
                      Figure 2.9. Kajiado district recorded the most persist drought in 1983-1985 with devastating
                                                        effect on the environment.

The NDVI patterns in Narok (Fig. 2.10) slightly follow that of Kajiado although the drought of
1982-1984 was less severe than in Kajiado. The severest drought was during 1999-2000.
                      3


                      2


                      1
     NDV An om al s
                 ie




                      0
        I




                      -1


                      -2


                      -3


                      -4
                                                                                  Jan - 87




                                                                                                                                                                                                        Jan - 9 7
                                                                                             Jan - 88




                                                                                                                                                                                                                    Jan - 9 8
                           Jan - 82

                                      Jan - 83

                                                 Jan - 84

                                                            Jan - 85

                                                                       Jan - 86




                                                                                                                               Jan - 9 1

                                                                                                                                           Jan - 9 2

                                                                                                                                                        Jan - 9 3

                                                                                                                                                                    Jan - 9 4

                                                                                                                                                                                Jan - 9 5

                                                                                                                                                                                            Jan - 9 6




                                                                                                                                                                                                                                                        Jan - 0 1

                                                                                                                                                                                                                                                                    Jan - 0 2

                                                                                                                                                                                                                                                                                Jan - 0 3

                                                                                                                                                                                                                                                                                            Jan - 0 4
                                                                                                                   Jan - 9 0




                                                                                                                                                                                                                                            Jan - 0 0
                                                                                                        Jan - 89




                                                                                                                                                                                                                                Jan - 9 9




                                                                                                                                                       Pe riod

                                 Fig. 2.10. The worst drought was experienced in Narok district in the year 2000.

The drought indices pattern in Laikipia district (Fig. 2.11) show there was a more elongated
drought between 1983 and 1986 compared to the drought episode in 2000. The severity of the
impacts indicates that coping mechanisms have broken down due to land use changes and land
degradation, among other factors. The situation was aggravated by migration of livestock from
neighboring districts, increasing competition for forage and water and also resulting in
conflicts with farmers and ranchers in Laikipia districts. The NDVI profile shows that drought
episodes in Laikipia district occurred in 1984, 1986, 1991-1992, 1994.


                            4                                                                                                                          43

                            3
                                           Fig. 2.11. The drought patterns in Laikipia district is similar to that of Isiolo district,
                                                   although the former experienced mild droughts compared to the later

The Isiolo NDVI pattern (fig. 2.12) indicates that the period between 1991 and 1992 was the
worst drought in terms of indices values. The main impact of the drought was on livestock and
wildlife. There were also drought occurrences in 1993, 1994, 1996, and 1997.



                                      4

                                      3

                                      2
                     NDV An om al s
                                 ie




                                      1

                                      0
                        I




                                      -1

                                      -2

                                      -3
                                                                                                  Jan - 87




                                                                                                                                                                                                                        Jan - 9 7
                                                                                                             Jan - 88




                                                                                                                                                                                                                                    Jan - 9 8
                                           Jan - 82

                                                      Jan - 83

                                                                 Jan - 84

                                                                            Jan - 85

                                                                                       Jan - 86




                                                                                                                                               Jan - 9 1

                                                                                                                                                           Jan - 9 2

                                                                                                                                                                        Jan - 9 3

                                                                                                                                                                                    Jan - 9 4

                                                                                                                                                                                                Jan - 9 5

                                                                                                                                                                                                            Jan - 9 6




                                                                                                                                                                                                                                                                        Jan - 0 1

                                                                                                                                                                                                                                                                                    Jan - 0 2

                                                                                                                                                                                                                                                                                                Jan - 0 3

                                                                                                                                                                                                                                                                                                            Jan - 0 4
                                                                                                                                   Jan - 9 0




                                                                                                                                                                                                                                                            Jan - 0 0
                                                                                                                        Jan - 89




                                                                                                                                                                                                                                                Jan - 9 9




                                                                                                                                                                       Pe riod

                                                 Fig. 2.12. Isiolo district experienced five major droughts between 1982 and 2004.


In the high potential districts of Nakuru (Fig. 2.13) the notable drought episodes occurred in
1984, 1991-1992, 1993-1994 with the severest occurring 1999-2000. This being an agricultural
district the effects was devastating on crops.
                3


                2


                1
DV An om al s
           ie




                0                                                                                                                                          44
  I




                -1
        Fig. 2.13. The NDVI anomalies for Nakuru district showing four severe droughts recorded for
                                              the district.


Murang’a district also recorded a number of drought occurrences with 1983-1985 being very
prolonged though the worst drought occurred in 1999-2000 (Fig. 2.14). The district has
experienced other minor droughts in 1986, 1992 and 1996. Generally, the high potential areas
have increasingly been affected due to other factors such as increased poverty levels
occasioned by breakdown of cash crop farming such as coffee and land degradation.
                  3


                  2


                  1
 NDV An om al s
             ie




                  0
    I




                  -1


                  -2


                  -3
                                                                              Jan - 87




                                                                                                                                                                                                   Jan - 9 7
                                                                                         Jan - 88




                                                                                                                                                                                                               Jan - 9 8
                       Jan - 82

                                  Jan - 83

                                             Jan - 84

                                                        Jan - 85

                                                                   Jan - 86




                                                                                                                           Jan - 9 1

                                                                                                                                       Jan - 9 2

                                                                                                                                                   Jan - 9 3

                                                                                                                                                               Jan - 9 4

                                                                                                                                                                           Jan - 9 5

                                                                                                                                                                                       Jan - 9 6




                                                                                                                                                                                                                                                   Jan - 0 1

                                                                                                                                                                                                                                                               Jan - 0 2

                                                                                                                                                                                                                                                                           Jan - 0 3

                                                                                                                                                                                                                                                                                       Jan - 0 4
                                                                                                               Jan - 9 0




                                                                                                                                                                                                                                       Jan - 0 0
                                                                                                    Jan - 89




                                                                                                                                                                                                                           Jan - 9 9




                                                                                                                                                   Pe riod

Fig. 2.14. NDVI anomalies for Murang’a district, note the persistent drought recorded in 1982-1985
periods. Yet the impacts of recent drought have been more devastating due other factors.
2.7. DISCUSSION AND RECOMMENDATIONS
The results from qualitative and quantitative analysis of NDVI show that it can be used as an
effective tool to understand past drought episodes. For example, the 1998, El Nino rains show
favorable agro-climatic conditions all over the country including the rangelands. This contrasts
sharply with the 2000 drought where the vegetation conditions in most parts of the country
were badly affected. The results indicate that the NDVI indices can be used to monitor
climatic conditions. NDVI has been proved to correlate with rainfall received although there is
a time lag of one to two months (it has been shown that it takes that period for vegetation to

                                                                                                                                                   45
respond to rainfall) between rainfall and NDVI received for an area. NDVI has also been used
effectively as a proxy to monitor net productivity of various ecosystems. DRSRS has been
using NDVI to monitor vegetation condition and primary biomass production in the Kenya
rangelands since 1993. This has formed a good database for future monitoring of trends in
drought patterns.


The qualitative NDVI products archived for a long period can be used to monitor the
progression of a given season. NDVI products for different years (Fig. 2.6) can be used for
early warning. By following the progression of a wet season on a decade basis any deviation
from normal is easily detected, as was the case in 1996. This drought was occasioned by the
failure of long rains in 1996 and was well captured by the NDVI products for that period.
NDVI can be supplemented with other information to strengthen the early warning component
to mitigate the devastating effects of future droughts.


Vegetation indices from satellite data can be used effectively to monitor drought. NDVI data
is available on a continuous basis and can be used in monitoring drought on near real time
basis as well as in trend analysis. NDVI data can be applied in both quantitative and qualitative
analysis of drought. The data can be used to assess drought duration, intensity and spatial
distribution of drought conditions. Satellite data is geo-referenced and can be captured as raster
data layer in GIS analysis for use in drought management.


Coordination between institutions dealing with early warning needs to be strengthened; these
include The Meteorological Department, Department of Resource Surveys and Remote
Sensing, the Drought Monitoring Centre. In addition, the crop-forecasting committee
comprising the Ministry of Agriculture, the Department of Resource Surveys and Remote
Sensing, and the Office of the President should be reactivated and supported. Finally the
linkages between early warning institutions and drought mitigation programmes run by the
government and NGO’s including WFP should be strengthened.
                                          CHAPTER 3
                    POPULATION GROWTH AND FOOD SECURITY

3.0. INTRODUCTION
DRSRS and the Central Bureau of Statistics (CBS) data were used in this chapter to assess
human population trends and food production. The temporal spatial population expansion has
led people to migrate into ASAL hence increasing the number of people vulnerable to drought
impacts. The results further indicate that whereas the area under crop cultivation has increased
significantly over time, crop yields have declined drastically. The results of the analysis
indicate considerable geographic variation in the distribution of well being in Kenya. The high

                                                46
potential areas exhibit significant variability of poverty levels. The poverty drought
relationship is positively co-related with ASAL having a high number of people affected.
        The ASAL were worst hit by drought compared to the high potential areas due to its
susceptibility resulting from natural climatic pattern. It was further observed that the number of
people affected by drought is on the increase and food security will remain a challenge to the
nation. Livelihood systems do not recover adequately to withstand the next drought. As a
result, any small shock such as a prolonged dry spell has a much bigger impact on people’s
livelihood than in the past. Rising poverty levels and recurrence of droughts exacerbates this
situation.


3.1. THE IMPACTS OF 2000 DROUGHT
The 2000 drought was declared a national disaster by the government. In response to the
government appeal, WFP spent US$ 102 million on food relief alone (WFP, 2001). In addition,
WFP required more than 15,000 tons of fortified blended foods for supplementary feeding
programmes in 11 of the worst-hit drought-affected districts located in pastoral, agro-pastoral
and marginal agricultural areas of the Rift Valley, North Eastern, Eastern and Coast provinces.
On its part, the Government spent in excess of KShs. 10.5 billion on relief food to combat the
drought emergency during the 2000-2001 financial years. Imbamba (2004), reported that the
La Nina drought (1999-2001) cost at least KShs 220 billion compared to the El Nino floods
which cost the country approximately KShs 70 billion. The year 2000 drought was the worst in
40 years.




3.2. HUMAN POPULATION
The high growth rate recorded for 1948–1962 for Kenya was due to the improved
completeness of the census, but there is no reason to doubt the rapid expansion of the
population recorded from 1962 onward (Table 3.1).


                      Table 3.1. Population size and growth rates
                                          Population (thousands)
                                                                       Growth Rate (%)
                        Census Year      Males     Females     Total
                            1948         2,680      2,726      5,406          -
                            1962         4,277      4,359      8,636         3.34
                            1969         5,482      5,460     10,943         3.38
                            1979         7,607      7,720     15,327         3.37
                            1989        10,628     10,815     21,443         3.34
                            1999        14,206     14,481     28,687         2.90
                         Source: CBS


                                                     47
                              Changes in population demography (1948-1999)

                 35                                                                 3.50
.




               Figure 3.1: Changes in human population demography in Kenya
                            between 1948 and 1999. Source: CBS

Figure 3.1 presents the underlying demographic changes of human population numbers
compared to inter-census population growth rates at the national level from 1948 to 1999.
There is a rising trend in the rate of population growth from the 1940s reaching a peak in the
1970s and a gradual decline thereafter. The first indications of a rising trend in population
growth in the 1960s undoubtedly helped spur the adoption of a national population policy and
programme in 1972. Similarly, the later estimates of natural increase as high as 3.38 percent
led to a renewed effort culminating in the creation of the National Council on Population and
Development, and numerous new initiatives in the 1980s. This has led to the drastic drop in
population growth rate.
3.2.1. Drought and Human Population
In 2000, Kenya suffered the third serious drought in ten years, with the ASAL being the worst
affected. During each succeeding drought, the number of people requiring emergency
assistance has drastically increased. The government estimated that 4.7 million people in 32
districts required assistance in 2000 drought. However, the World Food Programme assisted
3.3 million people affected by drought (WFP, 2001). About 25-30% of children under 5 were
estimated to be severely malnourished due to food scarcity during the drought in ASAL (Save
Children, 2000). WFP further observed that even if the drought situation subsided, families still
needed assistance into 2001 to enable them restore their livelihoods.
       Kenya’s population has more than tripled in the last three decades. The population has
not only increased in absolute numbers but also spatially (Figure 3.1). Situma (2003) reported
that spatial human population has up-surged in agro-ecological zone IV and V.         Migration
from more densely populated high-potential areas puts extra pressure on existing limited
resources in the ASAL. According to the 1999 human population census report, an estimated
12 million people now live in the ASAL districts compared to 8 million a decade earlier. This
constitutes about 36% of the country’s population. Of these, an estimated 20% live in the arid
                                             48
districts. The density of the population varies (Figure 3.2). Significant changes are taking
place over time, most notably in districts that are 50-85% ASAL. This is mainly due to
immigration. For example, Machakos district’s population density has increased over the last
20 years from 0-50 persons/km2 to an estimated 101-200 persons/km2. This has implications on
population’s resilience to drought impacts; the severity of drought impacts is increasing over
time. Increased population density over the years has led to land degradation, making it less
productive and more susceptible to impacts from extreme weather events.




                                             49
Figure 3.2 Spatial distribution of human population density between 1960s and 1990s. The spread of
population from the high potential areas (blue color) into the ASAL areas has caused high
environmental degradation (Source: CBS & Population Census)

3.2.2. Population expansion in the ASAL areas
The distribution of Kenya’s human population is to a large extent limited by geo-biophysical
characteristics, such as climate conditions, freshwater availability, infrastructure and
urbanization. The human population and associated economic development have been
increasing rapidly since 1960. Large areas, which were previously unsettled natural land, have
since been encroached by large human population and agricultural activities, which may not
only lead to severe land degradation, but also to the loss of large animal numbers and
vegetation.
       Rapid population growth in 1970s resulted in disproportionate rural-urban migration.
Over-concentration of services and administrative functions in the urban centers attracted
people in search of better livelihood opportunities. There was limited opening up of natural
landscapes for farming and therefore less land degradation. The high population pressure in the
high potential areas led to the migration of people into the medium and low potential areas of
Kenya. The nucleus settlements, previously, limited to different cultural values, political
inclination, and economic systems were disregarded. Accordingly, by 1990’s the distribution
of people was widely spread in the country, with high potential areas and urban centers densely
populated.


3.2.3. Human activity and drought
Human population in the arid, semi-arid and dry sub-humid areas is affected by poverty, thinly
scattered and with minimal means of ground communication. Information on drought impacts
often reaches the authority too late when the damage is already done. Interventions are equally
difficult to undertake to save lives. The survival of most people is therefore dependent on
weather patterns during the cropping seasons, cultivated area and inputs.         With the rapid
increase of population, these subsistence farmers continue to open more land for cultivation of
crops, thus exposing the soils to weather elements, over exploitation and eventual degradation.
Extended severe droughts normally leave populations hungry and poorer having lost livestock
                                               50
and crops and disrupted livelihood systems. This reinforces their poverty, which means they
cannot purchase and use optimal agricultural inputs in the next season so as to increase
productivity. This leads to increased poverty and the opening of more land for cultivation.
These trends need to be monitored: trends in absolute cultivated area and geographical spread.
Pastoralism is also still widespread in many parts of the Kenya rangeland. Probably this is the
oldest land use system still persistent. When the range was still expansive, nomadism as a
survival strategy in time of drought, used to save the stock. However, livestock movements in
search of pasture are much more restricted now. The pastoralists have therefore become more
vulnerable to droughts than before. The traditional dry season critical resource zones have
been lost to sedentary communities who depend on rainfed agriculture.
       There have been high population growth rates and densities in the dry sub-humid areas
as a result of migrants.      They more often than not put their new settlements under
environmental stress and therefore under the mercy of the climatic condition of these zones.
Refugee influx coupled with human population encroachments into the fragile dry lands that
has low resilience are major concerns to resource managers, especially in respect to the
survival of mankind in the long-term. At the same time, environmental refugees especially due
to several droughts in arid and semi-arid zones also make the dry sub-humid zones vulnerable
to degradation. Traditional drought and famine coping strategies are rapidly being lost.


3.2.4. Poverty Prevalence
Levels of poverty have been growing in most parts of the country for various reasons; low
economic growth, high unemployment rates, and breakdown of cash crop in the high potential
areas such as coffee farming. This partly explains why drought impacts are becoming more
severe, high potential areas that were in the past mildly impacted by drought were particularly
affected by the 2000 drought because of increased poverty. In the ASAL areas, livestock
marketing infrastructure is inadequate. It was observed during interviews that pastoralists are
now more willing to sell their herds during drought. Proper marketing infrastructure will
improve their economy and enable them to cope with the impacts of drought. In Kajiado
district, it was noted that cattle prices were one tenth of the normal prices during the drought
and this left the pastoralists poorer than they would be if proper marketing were in place.
       There is considerable geographic variation in the distribution of well being in Kenya.
Areas in the high potential zones exhibit significant variability of poverty levels (Figure 3.3).
ASAL districts show less spatial heterogeneity in poverty levels. Poverty densities are also
scattered indiscriminately across the country. An overlay of spatial data of poverty density and
drought vulnerability indicates that 85% of the people in ASAL are susceptible to drought
impacts. Some areas are endemic to poverty as well as drought.


                                               51
Figure 3.3: Poverty analysis in Kenya. Proportion of population below the rural poverty line (left), and
            poverty density (persons per Km2) (right): Source: CBS/ILRI, poverty mapping
3.4. AGRICULTURE
Droughts have a direct impact on food security. The most important cereal crops in Kenya are
maize, wheat and to a lesser extent rice, the others include legumes, sorghum and millet.
However, the latter are nationally insignificant in terms of production and consumption levels.
Maize and wheat are the staple food crops and are produced in the agriculturally high potential
areas (which comprise about 18% of the country's land area) under rainfed farming systems while
rice is mostly produced under irrigation. The grains reach all consumers in the country from
surplus producing areas to deficit areas through distribution networks.
       Guarantee of production and/or availability of the two staple foods to the consumers are a
major responsibility of the Government. However, the final level or net production is subject to
various factors including:
       Area put under the crop by farmers
       Prevailing weather and climate
       Soil fertility and water holding capacity
       Type of farming system
       Availability of inputs and their costs
       Cost of labour, land preparation implements, fertilizers, certified seeds, pesticides,
       herbicides, etc.
       Pre-harvest and post-harvest losses.
       Market forces


3.4.2. Maize and Wheat Crop Forecasting
The Department of Resource Surveys and Remote Sensing (DRSRS) has developed an elaborate
method of forecasting the production of wheat and maize crops. The programme estimates area
planted with maize and wheat, as well as estimated crop yield. The eventual outcome is crop
                                             52
production. Estimating demand and consumption of these crops is important aspect of monitoring
food security. Information on national food security status is necessary in the light of the
Government's long-term policy of self-sufficiency in feeding its people and maintenance of
strategic reserves. Area planted is estimated using sample vertical aerial photographs taken along
regularly spaced transects. Yield requires assessing of the healthiness of the green crop biomass
in the field using airborne radiometers. The biomass measurements are then converted into yields
through a field derived relational formula (Peden and Mwendwa, 1984). The estimated area under
the crop and the estimated yield for any area are used to forecast total crop production (biological
production or yield). The accumulated annual statistical information on these crop forecasts has
become an important component of the national data bank, which can be used, in strategic
planning. This data bank has previously been used effectively in national food crisis management
and the preparation of long term strategic plans. The quality and validity of these data are of
paramount importance and must be maintained in order to achieve the objectives of the
programme i.e. crop forecasting and creating an accurate database for national food security
planning.
       Due to the large area which is covered, multi-stage sampling techniques using satellite
remote sensing, systematic reconnaissance flights, and small format photographs and ground
sampling are used to determine the cultivation or the maize growing stratum in the country. The
techniques used often complement each other, but is no single technique that is wholly used.
Usually an area with 10% - 15% would determine the cultivation boundary, (Pilotto, 1988). The
procedures used include: satellite images, reconnaissance flights, and vertical aerial photographs
(Ottichilo and Sinange, 1987, 1988, 1990 and 1991).

3.4.3. Historical Trends in Maize Production
Crop data analysis shows that trends in area under maize has fluctuated between 0.87 million ha
and1.4 million ha (Figure 3.5). In a given season, the area cultivated is dependent upon family
subsistence needs, economic opportunities perceived and weather characteristics. Farmer’s always
plant maize crop each year regardless of risks anticipated, since it’s their staple food. Thus, there
is insignificant difference in area under crop over the years including drought periods. The gradual
increase in area under crop is due to expansion of agriculture in ASAL areas as well as along
riverbanks and other wetlands (Figure 3.4).




                                                 53
           Fig. 3.4. Encroachments by agricultural in the drier parts of Kajiado district.


Maize production has fluctuated over the years (Figure 3.6). Deficits were experienced in 1987
and all the years after 1989. The deficits are attributed to unfavorable weather (unpredictable
rainfall) over the years, losses due to drought, reducing soil fertility particularly in high potential
areas and poor agronomic practices often due to increased poverty levels. Other factors include
limited access to inputs especially fertilizer and certified seed; poor harvesting and storage
practices; unproductive cultural practices and pre-harvest and post harvest losses.
        The prospects of crop failure have been high in most parts of the country due to the onset
of drought before crop maturity. The country attained the lowest possible level of production in
1994 since 1989 due to prolonged drought. It was also probably because of reduction in area
under maize in the long rains season. The impact of dry spell of the 1993 was severely felt in
1994. Food deficits experienced in 1996-97 (Figure 3.6), were due to failure of the short rains. It
was estimated by FEWSNET (1997) that the failure of rains during the short rain season
contributed a 32% deficit on a national scale.           In 1996-97, GOK declared a shortfall of
approximately 8.5 million 90kg bags (761000 MT) maize prompting emergency plans for food
importation, increased coverage of relief distribution and waiving duties on all relief food stuffs.
        The average annual net maize production in the 1990s was 26 million bags (2.31
million MT) while maize consumption was 29 million bags (2.65 million MT). The national
population then was 27 million with a mean consumption rate of 1.2 bags (100kg) per person
per year (DRSRS database; FEWS, 1997). Thus consumption has been increasing due to
increase in population. Following the liberalization of the agricultural crop markets, the
purchase of food crops by the National Cereal and Produce Board (NCPB) has fallen
drastically.   Maize farmers prefer to sell their crop to private companies or individuals who
pay on delivery as opposed to delayed payment by NCPB.




                                                   54
                            130 0

                            120 0

                            110 0
  Are a (H a) in '0 0 0 '




                            10 0 0

                             9 00


                             80 0

                             70 0                                                     Are a unde r m aiz e
                                                                                       ine
                                                                                      L ar (Are a unde r m aiz e )
                             60 0
                                                          19 87




                                                                                                                                                                                                 19 9 7
                                                                   19 88
                                        19 85


                                                  19 86




                                                                                                           19 9 1


                                                                                                                             19 9 2


                                                                                                                                               19 9 3


                                                                                                                                                                 19 9 4


                                                                                                                                                                                   19 9 6
                                                                                         19 9 0
                                                                             19 89




                                                                                         Ye ars


                                Figure 3.5. Trends in area under maize in the country from 1985 to 1997
                                                       Source: DRSRS database


                             30 0 0


                             250 0
X 10 0 0 m e t Ton e s




                             20 0 0
              ric




                             150 0


                             10 0 0                                                                                  ic
                                                                                                            Dom e s t production
                                                                                                               ion
                                                                                                            Nat al cons um ption

                               50 0
                                                          19 87




                                                                                                                                                                                                     19 9 7
                                                                  19 88
                                          19 85

                                                  19 86




                                                                                                  19 9 1

                                                                                                                    19 9 2

                                                                                                                                      19 9 3

                                                                                                                                                        19 9 4

                                                                                                                                                                          19 9 5

                                                                                                                                                                                        19 9 6
                                                                                     19 9 0
                                                                           19 89




                                                                                              Ye ar


                                      Figure 3.6. Trends in per capita Maize production and consumption
                                       ∗ Assuming consumption of 98 Kg/capita/year Source: DRSRS




                                                                                              55
3.6. Expansion of Agriculture in Drought Vulnerable Areas
Expansion of agriculture from 1970s to 1990s is shown in figure 3.7. Agriculture expanded
largely into more vulnerable ASAL areas. The largest change in observed frequency of
agriculture expansion occurred in Narok, Machakos, Makueni, Kitui, Kajiado, Laikipia and
West Pokot districts.




Figure 3.7. Expansion of agriculture in ASALs, which includes the vulnerable areas between 1970s and
1990s. Source: Composite point maps from DRSRS aerial surveys data.

Moderate change in observed frequency of agriculture expansion occurred in Baringo, Taita
Taveta, Kwale, Kilifi, Tana River and Samburu districts (zones 4 and 5). Limited changes in

                                                56
observed frequency of agriculture expansion occurred in Marsabit, Moyale, Isiolo, Lamu and
parts of Garissa (zone 6). Most of the agricultural activities occurred in the dry sub-humid
zone and semi-arid zones. In recent times, agricultural activities are spatially encroaching into
semi-arid landscapes, but minimal activity was observed in arid areas. The proportion of land
occupied by agriculture increased tremendously between 1970s and 1990s (figure 3.8). These
ASAL areas are traditionally grazing areas for both wildlife and livestock. As migrants moved
into the vulnerable areas, they naturally settled on the wetter places, around water sources,
swamps etc which are vital locations for pastoralists during drought.

                              50
                              45
                              40                                                     Agro-Ecological Zones
        % En croach m e n t




                              35                                                     Z on e 1- H u m id
                              30                                                     Z on e 2 - Sub - H u m id
                              25                                                     Z on e 3 - Se m i- H u m id
                              20                                                                                  o
                                                                                     Z on e 4 - Se m i- H u m id t Se m i- Arid
                              15                                                     Z on e 5 - Se m i- Arid
                              10                                                     Z on e 6 - Arid
                               5                                                                 e
                                                                                     Z on e 7 - V ry Arid
                               0
                                   1   2        3        4        5         6    7
                                           A g ro - c lim a tic z o n e s




Figure 3.8. Expansion of agriculture in the seven agro-ecological zones between the 1970s and 1990s.
ASAL areas have had the larger proportion of agricultural expansion resulting in destruction of
ecosystems. Source DRSRS




Fig. 3.9. Wilted maize crop in Baringo district, 2000. Farmers had crop failure in three consecutive
years but were still hopeful of a harvest this time. Use of stream water for irrigation had been banned
due to water scarcity.

As the limited amount of water is diverted for irrigation (Fig.3.9) in ASAL areas, it becomes
more difficult for pastoralists to withstand drought impacts. Expansion of agriculture into
                                                                            57
ASAL areas is also causing destruction of vital ecosystems (Fig. 3.10), while this cultivation is
eventually not sustainable. These factors have lead to increased land degradation making
drought impacts more severe on the local residents. Such land use activities need to be
addressed at policy level when addressing drought impacts and its management.




    Fig. 3.10. Wetlands in Baringo district, which were traditionally used for dry season grazing by
                    pastoralists, are gradually encroached by agricultural activity.

3.5. AGRICULTURE IN NAROK DISTRICT (CASE STUDY)
Narok is one of the districts that have experienced an influx of human population from high potential
areas and a large expansion of agriculture in the recent past. In addition, the conversion of large areas
of land formerly used by wildlife and pastoral livestock into large-scale wheat farms has exacerbated
human wildlife conflicts and affected food production. Elephants in particular have increasingly
become a menace to subsistence farmers.
        The factors that are responsible for the low crop production in Narok include poor cultivation
techniques, lack of credit facilities and market outlets. However drought and crop damage by wildlife
remain the main constraints to rainfed cultivation. Droughts account for over 20% of crops lost before
maturity, whereas crop damage by wildlife accounts for 10% of the total production (DAO, Narok).
Crop damage by grazing livestock, destruction of seeds by birds (mainly the quelea birds (Quelea
quelea aethiopica), Russian aphids (mainly prevalent during drought periods) and other crop pests also
significantly accounts for the massive loss in crop production.




                                                   58
Fig. 3.11. Satellite imagery of 1975 and 1995 over the same landscape in Narok district showing
transformation of wildlife/ livestock grazing areas into wheat farms. The expansion of agricultural areas
has largely remained commercial and is of little benefit to the locals. Crop destruction caused by
displaced wildlife from their former range has increased food insecurity in these areas.

                Wheat Production (tons/ha) in Narok District, 1995 - 2000                              Maize production in Narok District, 1995 - 2000
 55,000                                                                               3    30 ,0 0 0                                                                          3.5

 54,000                                                                                                                                                                       3
                                                                                       5
                                                                                      2.   25,0 0 0
 53,000
                                                                                                                                                                              2.5
                                                                                      2    20 ,0 0 0
 52,000
                                                                                                                                                                              2
 51,000                                                                                5
                                                                                      1.   15,0 0 0
                                                                                                                                                                              1.5
 50,000
                                                                                      1    10 ,0 0 0
                                                                                                                                                                              1
 49 ,000
                                                                                      0.
                                                                                       5    5,0 0 0                                                                           0 .5
 48,000

 47,000                                                                               0           0                                                                            0
                                                                                                  19 9 4   19 9 5     19 9 6      19 9 7       19 9 8   19 9 9   20 0 0   20 0 1
       19 9 4      19 9 5   19 9 6        19 9 7          19 9 8   19 9 9   2000   2001
                                                                                                                                           Ye ar
                                                   Year
                                                                                                                    Are a (h a)                    P      ion ons
                                                                                                                                                    roduct (t /h a)
     a                       Area (h a)                      Product (t /h a)
                                                                    ion ons                b
Figure 3.12: (a) Wheat and (b) Maize Production in Narok District between 1995 and 2000. Source:
DAO, Narok.

Wheat occupies more than 50,000ha in Narok district and accounts for nearly 50% of the
country’s total production. The production of wheat declined from 2.7 tons/ha in 1995 to 0.9

                                                                                      59

								
To top